Iron-associated lipid peroxidation in Alzheimer's disease is increased in lipid rafts with decreased ferroptosis suppressors, tested by chelation in mice

Abstract

Introduction: Iron-mediated cell death (ferroptosis) is a proposed mechanism of Alzheimer's disease (AD) pathology. While iron is essential for basic biological functions, its reactivity generates oxidants which contribute to cell damage and death.

Methods: To further resolve mechanisms of iron-mediated toxicity in AD, we analyzed post mortem human brain and ApoEFAD mice.

Results: AD brains had decreased antioxidant enzymes, including those mediated by glutathione (GSH). Subcellular analyses of AD brains showed greater oxidative damage and lower antioxidant enzymes in lipid rafts, the site of amyloid processing, than in the non-raft membrane fraction. Apolipoprotein E ε4 carriers had lower lipid raft yield with greater membrane oxidation. The hypothesized role of iron in AD pathology was tested in ApoEFAD mice by iron chelation with deferoxamine, which decreased fibrillar amyloid and lipid peroxidation, together with increased GSH-mediated antioxidants.

Discussion: These novel molecular pathways highlight iron-mediated damage to lipid rafts during AD.

Highlghts: Alzheimer's disease (AD) brains have numerous markers for ferroptosis, including increased lipid peroxidation, reduced antioxidant levels, and increased iron storage. Lipid rafts in AD cases have increased oxidative damage and reduced antioxidant enzyme levels and activity which are most severe in apolipoprotein E ε4 carriers. Neuronal markers are correlated with lipid peroxidation, antioxidant defense, and iron signaling proteins suggesting that neuronal loss is linked to these events. Chelation of iron in the early-onset familial AD model reduces iron-mediated lipid peroxidation and fibrillar amyloid.

Keywords: 4‐hydroxy‐nonenal; amyloid; deferoxamine; early‐onset familial Alzheimer's disease; ferritin; ferroptosis suppressor protein 1; glutathione cysteine ligase modulator; glutathione peroxidase 4.

FIGURE 1

Oxidative damage (HNE, NT) to prefrontal cortex from age‐matched AD (n = 27) and cognitively normal controls (CTL; n = 25). (A) Schematic representing oxidative damage to lipids and proteins. Hydroxyl radicals (^.OH) are produced during oxidation of Fe²⁺ or Cu²⁺ (“Fenton chemistry”) and may oxidize adjacent proteins, lipids, or nucleic acids. HNE can alter proteins via Michael addition products through its double‐bonded carbons, or by forming Schiff bases through its aldehyde moiety. The bottom line shows the iron‐independent oxidative pathway to form peroxynitrite ⁻ONOO from O₂ ^•– and ^.NO, which can then oxidize tyrosine to NT. Aggregated human Aβ42 stimulated the formation of peroxynitrite which is neurotoxic in vitro. Non‐heme iron also mediates lipoxygenase activity. Dot blots are presented as percent change from control for RFUs for HNE presented as (B) cognitively normal versus AD, (C) APOE allele, (D) sex, and (E) NT. Significance, two‐tailed t test (B,C, E), one‐way analysis of variance with Tukey post hoc test (D): *P < 0.05, **P < 0.01, ***p < 0.001, ****p < 0.0001. AD, Alzheimer's disease; APOE, apolipoprotein E; HNE, HNE, 4‐hydroxy‐nonenal; NT, 3‐nitrotyrosine; RFU, relative fluorescent units.

FIGURE 2

Iron transport and metabolism in AD prefrontal cortex (n = 24) compared to cognitively normal age‐matched controls (n = 24). Percent change from control as RFUs from Western blots: (A) TF, (B) TfR/CD71, (C) DMT1, (D) FPN, (E) HCP1, (F) HMOX1, (G) HMOX2, (H) FTL, (I) FTH1. J, Correlation matrix for iron metabolism in clinical AD. Significance, two‐tailed t test (A‐I), Spearman correlation (J): *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. AD, Alzheimer's disease; DMT1, divalent metal transporter 1; FPN, ferroportin; FTH1, ferritin heavy chain 1; FTL, ferritin light chain; HCP1, heme carrier protein 1; HMOX, heme oxygenase; RFU, relative fluorescent unit; TF, transferrin; TfR/CD71, transferrin receptor.

FIGURE 3

Enzymes that repair or mitigate lipid peroxidation and facilitate GSH production. Percent change from control as RFUs for human prefrontal cortex western blots and enzymatic activity in whole tissue lysate for (A) GPx4, (B) GPx4 activity, (C) GPx1, (D) total GPx activity, (E) Prdx6, (F) FSP1, (G) GSTA4, and (H) ALDH2. I, Schema of enzymatic repair of lipid peroxidation. GSH cycle enzyme levels by western blots: (J) GCLM, (K) GCLC, (L) xCT/SLC7A11, (M) LAT1, (N) GSS, (O) GSR, (P) G6PD, and (Q) PGD. Significance, two‐tailed t test (A‐H, J‐Q): *p < 0.05, ****p < 0.0001. ALDH2, aldehyde dehydrogenase 2; FSP1, ferroptosis suppressor protein 1; G6PD, glucose‐6‐phosphate dehyrdogenase; GCLC, glutamate‐cysteine ligase catalytic subunit; GCLM, glutamate‐cysteine ligase modifier subunit; GSH, glutathione; GSTA4, glutathione S‐transferase alpha 4; GPx, glutathione peroxidase; GSR, glutathione reductase; GSS, glutathione synthetase; PGD, phosphogluconate dehydrogenase; LAT1, L‐type amino acid transporter 1; PRdx6, peroxiredoxin 6; RFU, relative fluorescent unit.

FIGURE 4

The lipid raft membrane fraction has increased damage and reduced antioxidant defense during AD. A, Schema of lipid raft damage, protective mechanisms, and cholesterol shuttling. Percent change from control as RFUs for western blot data and enzymatic activity (B) GPx4, (C) GPx4 activity, (D) GPx1, (E) GPx1 activity, (F) ALDH2, (G) APOE, (H) LRP1. Dot blots for (I) NT and HNE presented as (J) CTL versus AD, (K) APOE allele, and (L) sex. Significance, by two‐tailed t test (B‐K) or one‐way analysis of variance with Tukey post hoc (L): ^* p < 0.01, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. AD, Alxheimer's disease; ALDH2, aldehyde dehydrogenase 2; APOE, apolipoprotein E; CTL, normal control; GPx, glutathione peroxidase; HNE, 4‐hydroxy‐nonenal; LRP1, low‐density lipoprotein receptor‐related protein 1; NT, 3‐nitrotyrosine; RFU, relative fluorescent unit.

FIGURE 5

DFO chelation of EFAD mouse cortex for amyloid and iron. Percent change from control as RFUs for (A) insoluble Aβ fibrils, (B) soluble Aβ40, (C) soluble Aβ42, (D) HNE, (E) NT, (F) TF, (G) TfR, (H) DMT1, (I) FPN, (J) HCP1, (K) HMOX1, (L) FTL, and (M) FTH1. Significance, two‐tailed t test: ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. Aβ, amyloid beta; DFO, deferoxamine; DMT1, divalent metal transporter 1; EFAD, early‐onset familial Alzheimer's disease; FPN, ferroportin; FTH1, ferritin heavy chain 1; FTL, ferritin light chain; HCP1, heme carrier protein 1; HMOX, heme oxygenase; HNE, 4‐hydroxy‐nonenal; NT, 3‐nitrotyrosine; TF, transferrin; TfR, transferrin receptor.

FIGURE 6

DFO chelation of EFAD mouse cortex for antioxidant proteins. Percent change from control as RFUs or enzymatic activity for (A) GPx4, (B) GPx4 activity, (C) GPx1, (D) GPx1 activity, (E) Prdx6, (F) FSP1, (G) GSTA4, (H) ALDH2, (I) GCLM, (J) GCLC, (K) xCT. Significance, two‐tailed t test: ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. ALDH2, aldehyde dehydrogenase 2; DFO, deferoxamine; EFAD, early‐onset familial Alzheimer's disease; FSP1, ferroptosis suppressor protein 1; GCLC, glutamate‐cysteine ligase catalytic subunit; GCLM, glutamate‐cysteine ligase modifier subunit; GPx, glutathione peroxidase; GSTA4, glutathione S‐transferase alpha 4; PRdx6, peroxiredoxin 6; RFU, relative fluorescent unit.

FIGURE 7

Neuronal loss and transcriptional changes in AD prefrontal cortex. Western blots percent change from control as RFUs of neuronal markers by CTL versus AD, or Braak stage. A, Neuronal nuclei antigen (NeuN), (B) synaptophysin 1 (SYP), (C) postsynaptic density protein 95 (PSD95). D, Correlation matrix for variables with significant relationships to NeuN by CTL versus AD or Braak. Significance, two‐tailed t test (CTL vs. AD), one‐way analysis of variance (Braak) with Tukey post hoc, or Spearman correlation: ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. AD, Alzheimer's disease; CTL, normal control; RFU, relative fluorescent unit.

FIGURE 8

APOE allele associations with lipid rafts, antioxidant defense, and amyloid in the human prefrontal cortex: (A) NeuN, (B) LR yield, (C) LR cholesterol, (D) total LR protein, (E) LR APP, (F) total Aβ fibrils, (G) LR APOE, (H) LR LRP1, (I) FTL, (J) FPN, (K) LR HNE, (L) GCLM, (M) LR GPx1, (N) LR GPx1 activity, (O) LR GPx4, (P) LR GPx4 activity. Q, Correlation matrix by APOE allele. One‐way analysis of variance by Tukey post hoc, or Spearman correlation: ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. R, Model for ferroptosis in AD. Aβ, amyloid beta; AD, Alzheimer's disease; APOE, apolipoprotein E; APP, amyloid precursor protein; FPN, ferroportin; FTL, ferritin light chain; GCLM, glutamate‐cysteine ligase modifier subunit; GPx, glutathione peroxidase; HNE, 4‐hydroxy‐nonenal; LR, lipid raft; LRP1, low‐density lipoprotein receptor‐related protein 1.

EXTENDED DATA FIGURE 1

Oxidative damage (HNE, NT) in human prefrontal cortex and cerebellar proteins in age matched AD and cognitively normal control (CTL). Prefrontal cortex from Brodmann area 8, 9, or 10 and cerebellum were washed for measurements of (A) heme, (B) dot blot for 4HNE, (C) dot blot for NT, (D) HNE, (E) HNE, (F) total amyloid fibrils. Percent change from control as RFUs. Significance, two‐tailed t‐test (A‐D, F), one‐way analysis of variance with Tukey post hoc test (E): ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. AD, Alzheimer's disease; HNE, 4‐hydroxy‐nonenal; NT, 3‐nitrotyrosine; RFU, relative fluorescent unit.

EXTENDED DATA FIGURE 2

Iron transport and metabolism in AD versus cognitively normal prefrontal cortex and cerebellum. Western blots data as percent change from control as RFUs for (A) TF, (B) TfR/CD71, (C) DMT1, (D) FPN, (E), HCP1, (F) HMOX1, (G) HMOX2, (H) FTL, (I) FTH1. Significance, two‐tailed t test (A–I): ^* p < 0.05, ^** p < 0.01, ^**** p < 0.0001. AD, Alzheimer's disease; DMT1, divalent metal transporter 1; FPN, ferroportin; FTH1, ferritin heavy chain 1; FTL, ferritin light chain; HCP1, heme carrier protein 1; HMX, heme oxygenase; RFU, relative fluorescent unit; TF, transferrin; TfR/CD71, transferrin receptor.

EXTENDED DATA FIGURE 3

Enzymes that mitigate lipid peroxidation and facilitate GSH production. Western blot as percent change from control as RFUs or enzyme activity units for prefrontal cortex or cerebellum from whole tissue lysate for (A) GPx4, (B) GPx4 activity, (C) GPx1, (D) total GPx activity, (E) Prdx6, (F) FSP1, (G) GSTA4, and (H) ALDH2. I, Enzymatic repair of lipid peroxidation. GSH cycle enzyme levels by western blots, as RFUs: (J) GCLM, (K) GCLC, (L) xCT/SLC7A11, (M) LAT1, (N) GSS, (O) GSR, (P) G6PD, and (Q) PGD. Significance, two‐tailed t test (A–H,J–Q): ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. ALDH2, aldehyde dehydrogenase 2; FSP1, ferroptosis suppressor protein 1; G6PD, glucose‐6‐phosphate dehyrdogenase; GCLC, glutamate‐cysteine ligase catalytic subunit; GCLM, glutamate‐cysteine ligase modifier subunit; GPx, glutathione peroxidase; GSH, glutathione; GSR, glutathione reductase; GSS, glutathione synthetase; GSTA4, glutathione S‐transferase alpha 4; LAT1, L‐type amino acid transporter 1; PGD, phosphogluconate dehydrogenase; PRdx6, peroxiredoxin 6; RFU, relative fluorescent unit.

EXTENDED DATA FIGURE 4

The AD lipid raft has increased damage and reduced antioxidant defense. A, Schema of lipid raft damage, protective mechanisms, and cholesterol shuttling in AD prefrontal cortex and cerebellum. Western blots as percent change from control as RFUs or enzymatic activity: (B) GPx4, (C) GPx4 activity, (D) GPx1, (E) GPx1 activity, (F) ALDH2, (G) APOE, (H) LRP1. Dot blots for (I) NT and HNE presented as (J) CTL versus AD, (K) APOE allele, and sex for (L) cortex and (M) cerebellum. Significance, two‐tailed t test (A–J, N–S), one‐way analysis of variance with Tukey post hoc test (K,L): ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. AD, Alzheimer's disease; ALDH2, aldehyde dehydrogenase 2; APOE, apolipoprotein E; CTL, normal control; GPx, glutathione peroxidase; HNE, 4‐hydroxy‐nonenal; LRP1, low‐density lipoprotein receptor‐related protein 1; NT, 3‐nitrotyrosine; RFU, relative fluorescent unit.

EXTENDED DATA FIGURE 5

DFO chelation of EFAD mouse cortex comparing diet and IP for amyloid and iron proteins. Percent change from control as RFUs for (A) insoluble Aβ fibrils, (B) soluble Aβ40, (C) soluble Aβ42, (D) HNE, (E) NT, (F) TF, (G) TfR, (H) DMT1, (I) FPN, (J) HCP1, (K) HMOX1, (L) FTL, and (M) FTH1. Significance, two‐tailed t test: ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. Aβ, amyloid beta; DFO, deferoxamine; DMT1, divalent metal transporter 1; EFAD, early‐onset familial Alzheimer's disease; FPN, ferroportin; FTH1, ferritin heavy chain 1; FTL, ferritin light chain; HCP1, heme carrier protein 1; HMOX, heme oxygenase; HNE, 4‐hydroxy‐nonenal; IP, intraperitoneal injection; NT, 3‐nitrotyrosine; RFU, relative fluorescent unit; TF, transferrin; TfR/CD71, transferrin receptor.

EXTENDED DATA FIGURE 6

DFO chelation of EFAD mouse cortex comparing diet and IP for antioxidants. Percent change from control as RFUs for (A) GPx4, (B) GPx4 activity, (C) GPx1, (D) GPx1 activity, (E) Prdx6, (F) FSP1, (G) GSTA4, (H) ALDH2, (I) GCLM, (J) GCLC, (K) xCT. Significance, one‐way analysis of variance with Tukey post hoc test (A–K): ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. ALDH2, aldehyde dehydrogenase 2; DFO, deferoxamine; EFAD, early‐onset familial Alzheimer's disease; FSP1, ferroptosis suppressor protein 1; GCLC, glutamate‐cysteine ligase catalytic subunit; GCLM, glutamate‐cysteine ligase modifier subunit; GSTA4, glutathione S‐transferase alpha 4; PRdx6, peroxiredoxin 6; RFU, relative fluorescent unit.

EXTENDED DATA FIGURE 7

Neuronal loss in AD. Western blots as percent change from control RFUs of neuronal markers shown by clinical CTL versus AD or Braak stage. A, neuronal nuclei antigen (NeuN), (B) synaptophysin 1 (SYP), (C) postsynaptic density protein 95 (PSD95). Correlation matrixes for variables with significant relationships to NeuN for (D) prefrontal cortex and (E) cerebellum. Significance by two‐tailed t test (CTL vs. AD), one‐way analysis of variance (Braak) with Tukey post hoc, or by Spearman correlation: ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. AD, Alzheimer's disease; CTL, normal control; RFU, relative fluorescent unit.

EXTENDED DATA FIGURE 8

APOE allele differences in lipid rafts, antioxidant defense, and amyloid. Levels of (A) NeuN, (B) LR yield, (C) LR cholesterol, (D) total LR protein, (E) LR APP, (F) total Aβ fibrils, (G) LR APOE, (H) LR LRP1, (I) FTL, (J) FPN, (K) LR HNE, (L) GCLM, (M) LR GPx1, (N) LR GPx1 activity, (O) LR GPx4, and (P) LR GPx4 activity in prefrontal cortex and cerebellum of cognitively normal (open circles) and demented (closed circles). Correlation matrix of proteins analyzed by APOE allele in (Q) frontal cortex and (R) cerebellum. S, Schematic hypothesis linking oxidized lipid rafts to amyloid processing. One‐way analysis of variance with Tukey post hoc, or by Spearman correlation: ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. APOE, apolipoprotein E; APP, amyloid precursor protein; FPN, ferroportin; FTL, ferritin light chain; GCLM, glutamate‐cysteine ligase modifier subunit; GPx, glutathione peroxidase; HNE, 4‐hydroxy‐nonenal; LR, lipid raft; LRP1, low‐density lipoprotein receptor‐related protein 1.